Physiology of Pseudomonas aeruginosa in biofilms as revealed by transcriptome analysis

Abstract

Background: Transcriptome analysis was applied to characterize the physiological activities of Pseudomonas aeruginosa grown for three days in drip-flow biofilm reactors. Conventional applications of transcriptional profiling often compare two paired data sets that differ in a single experimentally controlled variable. In contrast this study obtained the transcriptome of a single biofilm state, ranked transcript signals to make the priorities of the population manifest, and compared rankings for a priori identified physiological marker genes between the biofilm and published data sets.

Results: Biofilms tolerated exposure to antibiotics, harbored steep oxygen concentration gradients, and exhibited stratified and heterogeneous spatial patterns of protein synthetic activity. Transcriptional profiling was performed and the signal intensity of each transcript was ranked to gain insight into the physiological state of the biofilm population. Similar rankings were obtained from data sets published in the GEO database http://www.ncbi.nlm.nih.gov/geo. By comparing the rank of genes selected as markers for particular physiological activities between the biofilm and comparator data sets, it was possible to infer qualitative features of the physiological state of the biofilm bacteria. These biofilms appeared, from their transcriptome, to be glucose nourished, iron replete, oxygen limited, and growing slowly or exhibiting stationary phase character. Genes associated with elaboration of type IV pili were strongly expressed in the biofilm. The biofilm population did not indicate oxidative stress, homoserine lactone mediated quorum sensing, or activation of efflux pumps. Using correlations with transcript ranks, the average specific growth rate of biofilm cells was estimated to be 0.08 h(-1).

Conclusions: Collectively these data underscore the oxygen-limited, slow-growing nature of the biofilm population and are consistent with antimicrobial tolerance due to low metabolic activity.

Figure 1

Oxygen concentrations in Pseudomonas aeruginosa biofilms. Panel A shows a representative oxygen concentration profile with depth in the biofilm. Zero on the x-axis corresponds to the biofilm-bulk fluid interface. Negative positions are located in the fluid film above the biofilm and positive positions are located inside the biomass. Panel B shows the coupling between oxygen and glucose utilization. The oxygen microelectrode was positioned at a location within the biofilm where the oxygen concentration was low. The medium flowing over the biofilm was switched between one containing glucose and ammonium ion (C, N) and a medium lacking these constituents (no C, N) as indicated by the arrows. The complete medium is present at time zero.

Figure 2

Spatial pattern of protein synthetic activity, as revealed by transient expression of an inducible GFP (green) in a P. aeruginosa biofilm grown in a drip-flow reactor. In this frozen section, the steel substratum was formerly at the bottom and the aerated nutrient medium at the top. Rhodamine B counterstaining (red) indicates the extent of the biofilm.

Figure 3

Comparison of transcript ranks for genes related to nutritional status and growth state. Shown are comparisons for selected genes involved in glucose uptake (A); oxygen limitation (B); iron limitation (C); presence of nitrate (D); and growth phase (E). Panel F shows the association between the difference in gene ranks for PA3622 (rpoS) and PA4853 (fis) and specific growth rate. Colored symbols correspond to individual data sets as given in Table 2 and Additional file 1. An asterisk next to a data point indicates a statistically significant difference between the indicated data set and the combined data of three standard comparator data sets (see Materials and Methods for specifics). In panel E, which concerns growth rate, the statistical comparison is to the two comparator data sets in exponential phase; the untreated control of Nalca et al was omitted. Where a label such as "Fe limitation" appears, it denotes a transcriptome that can be considered a positive control. Where no such label appears, a suitable positive control data set was lacking.

Figure 4

Discrimination of glucose metabolism (A) and homoserine lactone quorum sensing (B) based on differences in transcript ranks. Open symbols are glucose-absent or quorum sensing negative comparators in panels A and B, respectively. Filled symbols are glucose-present and quorum sensing positive comparators in panels A and B, respectively. Stars indicate drip-flow biofilm samples. The genes appearing in these graphs are annotated as: PA5564, gidB, glucose inhibited division protein B; PA3187, probable ATP-binding component of ABC transporter; PA2634, aceA, isocitrate lyase; PA3186, glucose/carbohydrate outer membrane porin OprB precursor; PA0485, conserved hypothetical protein; PA3724, lasB, elastase; PA3281, hypothetical protein; rhlA, rhamnosyltransferase chain A.

Figure 5

Comparison of transcript ranks for genes involved in stress responses and quorum sensing. Shown are comparisons for selected genes involved in oxidative stress (A); copper stress (B); efflux pumps (C); and homoserine lactone quorum sensing (D). Symbols correspond to individual data sets as given in Table 2 and Additional file 1. An asterisk next to a data point indicates a statistically significant difference between the indicated data set and the combined data of three standard comparator data sets (see Materials and Methods for specifics). Where a label such as "Cu stress" appears, it denotes a transcriptome that can be considered a positive control. Where no such label appears, a suitable positive control data set was lacking.

Figure 6

Comparison of transcript ranks for selected genes involved in synthesis of extracellular polysaccharides (A) and production of pili (B). Symbols correspond to individual data sets as given in Table 1. An asterisk next to a data point indicates a statistically significant difference between the indicated data set and the combined data of three standard comparator data sets (see Materials and Methods for specifics).